Apportionment of dietary diversity in wildlife

Research highlight: Apportionment of Dietary Diversity in Wildlife published in PNAS (Hannah Hoff et al.)

Led by Hannah Hoff, in collaboration with the National Park Service and the Brown University Herbarium, the lab just published a blockbuster paper that summarizes, critiques, and enhances how ecologists tend to talk about what wildlife eat. There are some surprising and potentially sensitive elements to the story—both for how we monitor and manage wildlife populations and for how we address our implicit biases when observing and reporting on wildlife as scientists.

Related content:

• Paper published in PNAS
• Hannah wins prestigious award
• Code to replicate analyses
• Results find attention in the press

​The two-sentence punchline: We combined two technologies—artificial intelligence and a genomics technique called DNA metabarcoding—to show that different species of large herbivores in Yellowstone have diets that are more diverse and complex than biologists had previously assumed. This is a big step toward understanding how so many species of large mammals can survive together in Yellowstone, and it suggests that maintaining plant diversity is a critical requirement for maintaining the diversity of migratory wildlife.

The slightly more elaborate version: The problem that this study addresses is a classic ecological paradox: “How do abundant herds of large mammal species divvy up the available plant resources to coexist?” Since these species seemingly rely on similar types of resources (grasses, trees), it’s not obvious why competition for food doesn’t cut the number of species down to just a couple of dominant competitors that eat different things.

These animals are difficult and dangerous to observe up-close, and it can be even harder to identify their food plants to the species level—even for expert botanists. But in spite of these challenges, biologists often try to characterize wild animal diets by watching them—straining to spot what they assume must be each species’ unique key to survival in a competitive world.

We wondered if wildlife biology might be shaped by stereotypes that get reinforced when well-meaning biologists report on behaviors that match their beliefs about what makes each species’ diet different. In this vein, the title of our paper pays homage to transformational work published in 1972 by Richard Lewontin, entitled The Apportionment of Human Diversity. This seemingly unrelated work on human genetics inspired us to think about wildlife in a new way. Lewontin demonstrated that most human genetic diversity exists within populations – it is not segregated between biologically different races in the way that some earlier geneticists had assumed – and this has become a cornerstone to our modern understanding of human genomics. Our paper offers a similar wake-up call for ecology: we consider whether and how wildlife biologists might reevaluate some of our preconceived notions about the biological differences between species in their environment.

We used dietary DNA metabarcoding and computer models based on simple AI techniques to help us figure out how many different ‘diet types’ exist within Yellowstone’s vast herds of wildlife – and whether each species has a unique diet type. What we show is that dietary differences between species ("niche partitioning”) are smaller than everyone had previously assumed – far less pronounced than similar research on African savannas. Instead, members of different species could have a lot of overlap in their diets and the amount of overlap depended on where and when they were feeding. Combining DNA technology with AI enabled us to break through long-standing stereotypes about what these animals eat, and look at the evidence in new ways.

Stories behind the science

• ​Two Studies, One Lesson: How Citizen Science is Transforming Our View of Yellowstone (Ecology Project International)
• An herbivore by any other name (The Wildlife Society)
• New research challenges animal dietary classifications in Yellowstone National Park (News@Brown)